INDEX

Explanations

words related to gaining advantage or improvement

oai_token-act-pair · gpt-3.5-turbo

phrases and concepts related to the idea of benefiting or gaining advantages

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 potato

-0.66

amb

-0.63

 bolt

-0.62

 Monstrous

-0.61

hus

-0.59

cluding

-0.59

 crisp

-0.58

pr

-0.58

 Traps

-0.56

 fluorescent

-0.56

POSITIVE LOGITS

benefit

1.01

iciary

0.94

ĸļ

0.90

doms

0.87

emale

0.82

 benefit

0.82

onies

0.80

 benefiting

0.77

igan

0.76

 financially

0.75

Activations Density 0.024%

words related to gaining advantage or improvement

phrases and concepts related to the idea of benefiting or gaining advantages

No Comments

No Known Activations